Neuronal electrical activity is the central underpinning of nervous system function. While understood as essential for over a century, the tools to study circuit level neurophysiology have remained largely unchanged in 50 years. The advent of molecular biology has dramatically advanced neurobiology by allowing molecular characterization of the nervous system but has not translated into significant gains in neural electrophysiology. Opto-molecular methods have revolutionized our study of neuronal connectivity, development, gene distribution, calcium signaling and recently, targeted neuronal activation (i.e. optogenetics). A glaring exception to this light-based revolution is the use of optical methods to monitor electrical activity. Intracellular calcium levels and metabolic signals are often used as a surrogate marker of electrical activity, however they are temporally delayed, do not detect subthreshold events and more often than not fail to capture the relevant suprathreshold activity. The PIs laboratories, as part of a multi laboratory collaboration have been developing genetically encoded voltage sensors based on fusions of green fluorescent protein orthologs and voltage sensing domains. Our grant members have published most of the significant advances in genetically-encoded voltage sensors in recent years. Our most recent probes, Arclight and ElectricPK significantly improved the signal size and response speed of fluorescent voltage probes. The current application will continue this successful collaborative search for voltage probes. We are seeking probes which combine large F/ V signal sizes, a range of useful response speeds and red-shifted fluorescence spectra. During this previous funded period time, we discovered that by altering the voltage sensor domain, the linker length, the fluorescent protein and by introducing point mutations in the fluorescent protein, we could develop probes with vastly superior signal size and response kinetics. We also confirmed, however, that a purely empirical step (i.e. large scale screening of single, incrementally-modified constructs) is required to make dramatic improvements in response properties. We will employ a staged evolution approach involving successive rounds of directed and random sequence modification followed by direct testing in mammalian cells. The current experiments will be an advance over all previous studies in two important ways: i) we will create vastly greater numbers (20x) of potential probe (thousands) using domain swapping and site directed / random mutagenesis and ii) the larger numbers of constructs will be prescreened by an automated, robotic microfluorimetry method which evaluates the fluorescence signal size and speed in electrically-active mammalian cells. Finally, all successful candidates will be validated for in vio functionality in Drosophila circadian neurons and rodent somatosensory/barrel cortex.